Presumptive evidence for a viral etiol gy of multiple sclerosis might come from the demonstration of a virus-induced disease in animals like multiple sclerosis. Experimental Theiler's virus infection of mice represents the most interesting animal model of virus-induced demyelinating disease currently available since, (1) there is persistent central nervous system (CNS) infection, (2) in the chronic phase of infection pathol gical involvement is limited to CNS white matter, (3) myelin destruction appears to be immune-mediated, and (4) the appearance of clinical disease due to demyelination occurs after a long incubation period. Adoptive immunization experiments to further document the immunological nature of the demyelinating lesion are planned, particularly with the intent that the immune response resulting in tissue injury is directed toward virus antigen. In chronic cytolytic virus infections where virus replication takes place in only a small percent of cells, immune-mediated mechanisms may greatly amplify tissue injury. However, the usual mechanisms of virus-induced immune-mediated cellular injury appear to have little relevance for non-budding, cytolytic viruses because of the absence of viral antigens on the cell surface. Additional studies will be conducted to determine if bystander killing occurs in the context of Theiler's virus infection in vitro. The mechanisms involved in the establishment and maintenance of persistently infected CNS cultures with Theiler's virus will also be examined.